, t rovP' I 'EZ

44 14

r Wviw SATEm

57~~~~~ U,r 4 44~4~

MINWR OF UPPL

R~$E 7Rc094

1PORT No. 0//4

0 0

no:Prpaatonan Pop rtisor id w, lcl

4 ~CA)PATn

41 Am Ses AWI

4

4< 4444<

I

44

4 k

<4 /

>4K

444

4 .4 S

4>

4;

I~ ,*

A.LE.?ublkWaa Scion

4<

44

'4

41 I -~

4>4

A

4'

A

4'44

~ K: ~< <4 4 4~>< q;<4 < 4

444/ ~4 44 ' 44<' 4<

4< 44 44

4 4

LIMCLASSIFIEDE.R.D.L. gAL249. J'L.R. -51OZ1.

tIdNISTRY OF SUPPLY.

L.ILO'JIVES RLSLR'CH -46 DLVLQOPI I*NT LSj ABLI$H'ELT.

L.R...U. RLPORT NO. 2//49.

The Preparation and Properties of DiethyleneGlycol Dinitrate.

Part III. Batch Plant Manufacture

by

S. E. Smith.

This re,port does not contain any infori,ation ofoverseas origin.

Submitt ed by--' ---

/ S.i. I.

Approved by

A /C.S./E.R.f.L.Waltham Abbey,

Essex.

Best Available Copy

UNCLA7- '..

U N2CLV DISTRIBUTION.

LXTERNALJ. T.P.A.3/T.I.B. (41).

Ministry of Suip-pl.

Chief Scientist.C.S. (M).C.I.

I.A. (Air) R.A.L., Farnborough.C.S.A.R. (2).A.D. Aim.P. (X).D.lvl.X.R.D. (3).D.O.F. (X).P.D.S.R. (D).Sec. O.B. (2).S.A.C. :(47)-'

War Office.

M.C. of S. Shrivenham (2).

Air Ministry.

D.D.O.R. ('2).

Adr,dralty.

C.I.N.O.Supt., R.N.P.F.

Consultant.

Dr. G. Rotter, C.B., C.B.E.

INTER1AL.C.S. ,E.R.D.E.S.P.R.I.S.P.R.I:.S.L.I.S.C.E.Mr. S.E. Smith.Tech. Reg. (2).Library 3).

Information Bureau (2 + stock)

UNCLASSIFIED

UNC"

IVMTISTRY OF SUPPLY

EXPLOSIVES RESEARM eM IEVqLCPivNT ESTABi--&Ev"NT

TBE PMAP.RATION AIMD PRWERTIES OF DIEnIYE

Parts I and II. E.R.D.E. Report No. 18/R/48 - Ref. C.R,To. 7/.U/2/1

BY

S. MasteranS.E.SmithW.E. Turner

Parts Ill. E.R.D.E. Report 2/R/49 - Rof. X.R. 510/1.IV. " " 3/R/49 - i X.R. 510/2V. " 4/R/49 - it X.R. 510/3

Vi. 2 2/R/49 - x.R. 510/5by

S.E.Smith

WALTAM ABBEYESSEX

L :!AS€,- J.

I. objects of the ixivesti,ation

To exa,-dne the properties of diethylene glycol dinitrate, and to

develop a .method for its :anufacture from diethylene glycol.

II. Scope of the investigLation

The require:,ients of commercial diethylene glycol for nitration to the

dinitrate for Service use have been investigated. The optimum conditions for

the nitration, and for-E- stabilisation,'f diet,ylehe -glydol :dini:tr ' have

lbeenJde-termi,ned .by laboratory -id se! i-etzhnal Usae 'linveSigatiols, and

-ath-nd; cohtinuous4Pil:6t plaants o s oaions ha e beer developed.

The -manufacture of D.E.G.1 and recovery of the spent acid have boen

established on the sei-manufacturing scale and quantities of the product

supplied for experimental propellant :ianufacture.

The cheistry of D.-.-J., the mechanism of its nitration and the

reactions of the spent acid have boon Livcn some fund :,untcl study.

No difficulty has boon met in obtaininL D.-.C-.-. of accuptablo purity

and chemical stability by these processes.

The maxi-aur overall moufacturing yild of D..1.C.1. so fL.r obtL.ined

is 90 per cent theory as against 94 ocr cant for nitroLlyc.;rimn. This is

a disadvantage of D.E.G.N.

The spent acid from D.].C-.7C. nitration retains some D. *. 1. in

solution and fues off at ordinry temperatures after a 'life duponding

on the co.position of the acid. An acid composition for continuous nitration

has buon devisud to give a spent acid with a safe lifo around 12 hours at

200C. A procedure has boon dovolopud in which th spunt a cid from th

continuous process is run directly don a dinitration toaer in which the

dissolv,d organic matter is d,stroycd. This involves the r"covury of N02as 60 per cent nitric acid and some additional nit- i c acid concentration.

The physical, cho.Acal, explosive and physioloLical properties of

dictylVlonc glycol dinitratc so obtaincd have boon dotor.ined.

Conclusions

1. D.E.G.T. can be safely =4nufacturcd in the same types of plant, batch

or continuous, as are used for nitroglycerine. The continuous process is

more suitable for operation with a continuous waste acid donitration. In

both batch and continuous nitration the safety of D.E.C-.N. enables

sinVlifications to be made. The yield of D.E.C-.N. is less and the

consumption of nitric acid is somewhat higher than for nitroglycerine.

2. The spent acid from D.E.G.N. nitration is unstable and needs to be

dccomosod directly and continuously. Donitration tower practice has boon

found suitable for this purpose and c-n bc safely linked up with continuous niti

nitration.

3. D.1.C.. is superior to nitroglycerine in safety in handling.

4. D.2.C-.IT. appars to have no disadvantageous plysiologic-l effucts

during the short manufacture to dat.; the lon, ter, &ff ct on hu-lth

of the o:crars is yet to b, dLtern-a.

CONTENTS.

I. Semi-technical scale plant.

1. Introduction.2. Description of Plant.3. Operation of Plant.4. Discussion and Results.

II. Batch Pilot Plant.

1. Introduction.2. Description of Plant.3. Operation of Plant.4. Cozmaents on the Design of the Plant.5. Plant Investigational Work.6. Yields and Flow Sheets.

Appendices.

Figures I. Layout of Semi-Technical Scale Plant.II. Photograph of Semi-Technical Scale

Plant.III. Photograph of Batch Pilot Plant -

Nitrator.IV. Photograph of Batch Pilot Plant -

Prewash Tank.

Drawings of Batch Pilot Plant.

1. Flow Sheet.2. General Arrangement.3. Layout of Piping.4. Mixed Acid Gauge Tank.5. D.E.G. Gauge Tank.6. Nitrator Drowning Valve.7. Nitrator Cover.8. Prewash Tank.9. Waste Acid Tank.

Reference X.R. 510/1.

I. Semi-technical scale plant.

1. Introduction.

A 20 lb. batch plant was brought into operation as soonas sufficient data relating to the nitration of D.E.G. wasavailable. Its object was to provide further informationabout batch nitration and also to meet an increasing demandfor DEGN. This semi-technical plant work therefore precededsomc of the work presented in Part II ,nd inforiation gaincdfrom it contributed to the conclusions rc ched at the end ofPart iI.

At 2 ltcr date, the 20 lb. seiai-tLchnical plLnt w-,s aginused, to compoar directly tht yields obttin:bli by the ,)rocessrccoi,1(ended in Prt II, viz. nitration with 2.57 p.rts of 72/23mixed acid, against the yields obtiin-blL by etthods which wercused in German factorics during the 1939-45 w,-r.

2. Description of Plant.

The layout is shown di'gr,mmatically in Fig.I, 'nd Fig.IIis a photogr.-ph of the pl'nt.

A miniature N.G. typ. nitr,,tor sc--r-tor w-s used, consist-ing of a 4 gallon lcad pot, 1 ins. di ,. by 10 ins. dep, witha single lced cooling coil, 7/8" dir,ri. 82 ins. long, a n inner-nd an outer ring of iir stirring coils on thL bottom, -,nd abottom outlet through which displ:ceLent acid ws introduced,,nd by which the contents could bc discharged, through.- stone-

wnr. cock, into a drowning tank. The top of the nitrrator hnd

a conical rim 2 ins. deep, with a sloping gutter leading to theprewash tank. An outlet near the top of the nitrator wasconnected to a fume extractor.

In addition to air stirring, a stainless steel impellorstirrer, driven by a geared electric motor, was also fitted.

The DEG was fed from a small tank mounted above the

nitrator. When air stirring was used, injection was by a mini-

aturc glycerine injector, but with iiechanical stirring, the DEG

was run in a thin strcai on to the surface of the acid into the

vortex created by the stirrer, from a coi,1po,_O_zle. Cooling was

by ice water, flowing by gravity from za tank fixed ct n sufficientheight.

A tank for displqcc-,icnt acid, with a bottow outlet closedby a cock, was mounted abovc the nitr-tor.

The drowning t'nk vas of 20 g allons c--ycity and had an nirstirring coil. It w-s noriailly. kept t-!o thirds full of watcr.

The prL-w-sh t-nk w's stainless steel, 10 ins. diam:etcr,

9 ins. deep, 2 gal. capacity, with sloping bottom, and - cool-

ing jacket through which w-tcr.coul6 be p-ssed. It ,a,is fitted

with a stainless stcel ;ir stirring coil, !7nd h-d -bottom run-

off to which ws attached a rubber tube. During v.-shing the end

of this tube w s held ,bove the levcl of the contints of the prc-

/warshcr,

washer, and a stream of air blovvn through it. The re-ashcrwas also fitted with a skiL.er, of the sa,ie design ns thnt

used in washing N.G., but sieller, and ian6e of stainless atoe l.

The skirmiiings vcre passed into 9 bucket or n sLttling t--nk.

The washer w,s i lcad trnk 15 ins. di,m. by 15 ins. deep,

capacity 5 gallons, with a sloping bottoii ,nd - 2 ins. deepconical rim at thL top. Stirring w-s by acans of -ir blo,ln

through a pcrforoted coil of compo tubing. It s fittcd ,itha lend N.G. pattern ski-miUr, nnd, bottoiI run-off carryingrubber tube which, during vi-shing, w-s kept -bovc thL lLvel ofthe contents, nnd connLcted to n ir supply. A 100 gillonpan was used rs a vish w tLr settling tank.

3. 0-pLrrition of Pl,ant.

In oreLr to carrr out - nitr,-tion, the mixed a,cid vv's firstweighcd out nnd plictd in th4 nitrator. ThLn enough mixed acid

was plnced in the displacement tank to fortify thL wstc acid to

be produced in the nitration: to this was ndded about 3 g:,llons

of old fortified w-stc acid (F. I.A.). ThL cooling vatcr wa,nsthen turned on to thc coils in the nitrator, ond the stirrer

started. A little DEC was placed in the DLG tqnk ;ind levclledoff on the sight glass to a fixed level, to provide a datum line.

The DEG necdcd for the nitration va-s thLn veighed out :-nd placed

in the DEG tank.

When the acid in the nitrvtor had cooled to 150C., addition

of DEG was conmenced, and the rate of inflow rLgul.ted so as to

kcp the temperature steady at 150C. throughout the nitration.

This usually mfant, in summer, n rptc of flow of cooling water of

5-6 gnilons/min. entering at 0-10 and issuing ot nbout 70 - 80C.

Under thesC conditions 13 lb. of DEG could bc nitr-atcd in an hour,

using 6-8 cwt. of ice. Addition of DEG was stopped whcn thG

level in the sight glass re,achLd the dzatuw line.

one minute ,ftcr all the DLG hrd bccn tdded, stirring \, s

stopped and the cdOitents of the nitra.tor llo.' 'd to scp-r,te.

This took about 10 Tinates, :nd its progress w,s follovied by

observing the gradual :vpc:)rnnce of the nitr:-tor coils through

the ester layer as it cleared. When sepf r-tion \ s complete,

sa plcs of c,-ch layer vcre taken by m ns of a pipcttc, for :,n-l-

ysis. Fortified waste acid was run in via th, bottom inlet

of the nitratar to lift the charge, so th,t thf lcvcl of the

cster layer rose and it Lventually ovLrflowed dow;;n thL sloping

gutter into the prc-wash tank YhcrL it ',s received in i-,tLr

gently stirred by air. . If the chrge wa's a 1)rg one, the

pre-VRsh ,as done in t,, prts, in which c sc displ-cmcnt took

about an hour from the end of nitrotion.

When all the ester had been displaced over into the pre-

wash, a little of the fortified waste acid was removed from the

bottom of the nitrator so as to drop the level slightly, and

the stirrer was startcd again to mix up the contents and stabi-

lise the waste acid. The rest of the contents wcrc then rc-

moved, weighed, and stored for fu*urc use as displacement acid.

From the weight if this acid, and that of the acid placed in

the displacement tank the weight of the wa-stC acid formed was

obtained by difference.

The acid ester laycr was washed .with about tvwa thirds of

/its

-2-

its weight of water, in the pre-wash tank, with cooling waterin the jacket to keep the temperature down to below 250C., thepre-wash liquors being recovered at a concentration of 30 percent or above. At the conclusion of the wash, after settling,the wash waters were skimmed off and weighed, any ester accident-ally removed and settling to the bottom being saved and returnedto the rest of the batch. The ester itself was run off fromthe bottom of the washer into a jug or bucket.

About 8 inches of alkaline wash waters werc placed in thewasher, and the air stirring started before the ester was addedfrom its jug or bucket. The pre-washcr was washcd with viatcrinto the jug or bucket, to rClIove all tracces of estcr, and thewashings added to thu alkaline wash.

The alkali wash i-.ratLrs wcre mcde up and stored in corboys,from which they wfrc measured into th , washers in gallon jugs.Tapviater from taps placed above the washers was used for the pre-wash and for the water washes.

At the end of each wash, after the two layers had had timeto separate and clarify, the aqueous layer was skinmaed off andrun into the wash water settling tank. More wash waters werethen added, and stirring restarted.

At the conclusion of the last water wash, after skimmingoff the water, the ester was run off from the bottom of thewasher through a rubber tube on to a large filter paper in a coni-cal funnel and filtered into alarge Buchner flqsk. It was thenweighed and stored in a carboy.. The filter popcr removed thedust and all suspended droplcts of water. On cooling the filteredester, however, more water came out of solution, nnd the esterbec3me milky, but it cleared again on warming.

At the conclusion of each nitration the DEGN which hadaccumulnted in the wash water sottling t.mnk wns rccovcred andweighed; it sometimes amountcd to 4 oz. out of a theoreticolyield of 24 lb., i.e. 1 per cent.

The details of .- ty-picnl nitration nrL "ppLnded :-

32 lb. of mixed acid of composition 73.39 per cent IR03,26.48 per cent H2S04 and 0.13 per cent H20 were weighed out andplaced in the nitrator.

10 lb. of the same mixed acid and 34 lb. 6 oz. of old forti-fied waste acid were weighed into the displaceiient tank.

After filling the lead etc., of the DEG tank up to the mark,

13 lb.0 oz. of DEG were placed in it.

Room temperature was 250C.

Stirring was started in the nitrator, the mechanical stirringbeing used. A slow streamn of air was blown down the bottom inlettube to prevent pockets of acid or DEGN forming in it, and thecooling was started.

At 9.50 a.m. the temperature of the nitrator had fallen to140C. and the cooling water was leaxing at 80C. at a rate of 6galls/min. The addition of DEG was commenced.

/The-3-

The following were the rccorded tt,&pcraturcs of the

nitrator and the cooling watcr*lcaving thc nitrator, and tharatc of flow of cooling water during the nitration :-

Tcmp, of Nitrator. Exit terap, of cool- Rntc of flow.

ing wptcr.

1400. 3.0C. 6 galls/rain.

170C. 8.80C. 6 '

170C. 9.30C. 6 " "

17.30C. 10.0 0 C. 6 "

15.50C. 10.50C. 6 "

All the DEG had been added by 11.03 a.m.

The nitrator was then allowed to cool to 12.50C., stirring

was stopped at 11.10 a.m. and the flow of cooling water reducedto keep the temperature steady. After ten minutes the ester

layer was almost water-clear. At 11.23 a.m. samples of wasteacid and ester layer were removed for &nalysis, and displnccmcnt

was commenced. The cooling woter had to be turned full on to

cool the displacement acid, which was at 240C. 58 lb. of acidwere recovered from the nitrator, and 6 oz. from the disploccmcnt

tank, so that the weight of the wastc - -id forTacd during the

nitrction wa7s 14 lb. 0 oz.

The composition of the wcstc acid u-,s 12.77 per ccnt., ofl0 3 ,53.93 per cent.. H2S04 , 29.59pLr cent h20, 3.41 per cent LGN.

The composition of the estcr lryer was 19.21 per cent 12O 3 ,

1.02 per cent H2 S0 4 , 77.27-per cent DLGN (2.5 :er cent H20 ctc.-

The prc-viosh w ", &onc in two p rts, for v,hich 4- inches and

43 inches of vutcr wcrc used respctively. The last of the cster

went into th. prc-wa sh ta-nk a,t 11.55 .ra. Nitr'tion nd scpor,tionhad thus t kcn 125 minutes.

The prc-w sh liquors werc maixed togethcr and v.igh,d. Theyweighcd 31 lb. 2 oz. -nd h,d the following composition :-

17.62 per cent HN03, 1.70 per cent H2SO4, 0.04 per cent H1N02,

1.48 per cent M -

The maximum tcmpcrnturm rcichcd in thL pre-v;sh vwas 260C.

After each prirt had rcCeived its pre-wash it was removed fromthe pre-washer in a jug nnd poured into the first olkaline wash

in the led viasher.

1st Alklinc Wash - 3 gallons of wiarm 2 per cent sodium bicarbonatesolution at 25'0C. were placcd in the vsher ,-nd the aiir stirring

wns started. The DlEGN from the prc-v,;sh w'>s added in tio parts,

follo,cd by the washings of the pre-viashLr and jug. This v,ash

was continued for 21 hours after addition of the second lot of

DEGN. The composition of the wash waters finally was DEGN 0.513per cent., NaN03 0.952 per cent., N2C03 C.933 per cent., NaN02

1.125 per cent./2nd Alkaline

-4-

2nd Alkaline iash - After se-paration of the first alkalinewash, another three gallons of 2 per cent sodiui. bicarbonatesolution was added. . The tcriperature of the contents of thewasher wos 250C., and the wash was continued for 3f hours.

1st W,,ater ',ash - For this, 3 gallons of water at 256C. wereused. The wesh was continued for 0 hours, pH = 9.41. Nonitrite could be ostia.-tcd in the wash waters but they gave ab [ue colour with storch and Kl.

2nid 1iottr 1;Vosh - 3 gallons of wctcr at 250C. for 3 hours, pH -8.47. Heat test of DEGN 4 mins.

3rd Water Wash - 3 gallons of water at 250C. for 2 hours.Heat tcst of DLGN = 7 mins.

4th Water Wash - 3 gallons of water at 2500. for 3 hours.Het test of DEGN = 3 mins.

5th Wnter Wash - 3 gallons of water at 250 C. for 2" hours.Heat test of DEGN = 13 mins.

The yield of DEGN was mode up as follows -

Recovered at end of washing process 20 lb. 12 oz.it fro,u s.ttling tank 3 oz.

Disposed of in s.xpl s for Onialysis 1 oz.

21 lb. 0 oz.

The thGoretically pos&ible yield is 24 lb. 1 oz. There-fore th(. perc,ntngu yield is 87.4 per cent.

ThL following ainAlysis sho s the distribution of them,tLriels involved in the nitrntion :-

B. 42Slb H20 f l + D=f Total1b.4 100 0b, 100 l0b1 .5 (Uo

Mixed acid 23.48 100.0 8.47 100 .042 0.95 13.0 31.99' Waste acid 1.79 1 7.63 7.56 89.3 4.18 93.8 0.38 1.6 1 13.91

Acid Ester 6.95 29.6 0.31 5,661 0.77 17.3 23.9 99.6 1.00Pre-wash water 5.47 23.3 0.53 6.2 [24.60) - 0.46 1.9j 31.11lst.ALk."Wash 0.23 0.981- - - 0.16 0.71 32.0As D I! 16.00 68.2 - - - 24.9 -i .. . t. . ... t .... . ....... .. .t. . . . .

The figure for the weight of the acid eEter above is

obtained by difference from the initial weight of P .A. andDEG and that of the waste acid. The latter is itself

obtained by difference, and therefore neither are very accurate.

The figure for the weight of the 1st Alkaline ,ash is calcu-lated from the total depth of the wash waters and DEGN, namely9 inches.

The above figures are a rough guide only on account of the

uncertainty necessarily attaching to any estimate of the weight

of the waste acid.

Towards the end of the series of nitrations, it was desired

to produce DGN in larger quantities, and to avoid having tospend two days on each 20 lb. batch, four nitrations were carriedout in one day and all four batches washed together in a 100

gallon pan. The nitrations worc perfored as beforc, but insteckof displacing the ester layor over into the pro-wash, the holecontents of the nitrator ware removed into a gloss separotorconsisting of an invcrted bell-jar with o glass cock ciLntcd

into its Wpex. Herc the nitration mixture wcs jllo'vd to sep-arote, the lower acid lnyor bcing run off into - bucket and theacid cstor layer being WIshed p-rtly in the rc-7rshor Wpartly in n stlainlss stocl bucket, nd tho Uttur hft!y rdsseparatcd in the inverted bell jnr.

The wistc acid w,s tSkLn outside cnd alloycd to fum(-off.This took novcr lLss thn thrc hours on hot suam"cr 6cy. SoOj-

tiMLS it took so long th,t it had to bc artificially fun(d-off byinjDctionb of live steam. Whilst one batch was separating, the

nitration of another butch was proceeded with.

The DEGN was collected together in a 100 gallon pan under

sodium carbonate 2 per cent solution, and then washed with the

same solution added to a depth equal to that of the DEGN, and

later with alkaline sulphitc solution and with water. The con-

tents of the pan were stirred by air at 12 lb. pressure introduced

through a coil of perforated compo tubing, this gave a thick

stiff creamy emulsion. It was found necessary in ordcr to pre-

vent the workcrs in the building from getting headachcs, to Vit

a cover to the pan and extract the fumes through a catchpot. As

much as I lb. of DEGN would be rccovrd from this cotchpot aftr

four or five washes of a 100 lb. batch of DEGN.

4. Discussion and Results.

In the first three nitrations, the nitrntor wrs air-stirred

and the DEG was introduced by maons of a miniature glycerine

injector. This was found to be much slower and to produce more

fumes than mechanical stirring. The latter circulated the con-

tents of the nitrotor over the cooling coils instucd of just

throwing it about, and thus gav more cfficient cooling, so th7t

the time of nitration wos reduced by 25Q.

Nitrations were usually cnrricd out qt 150C. A fcw were

done at higher tempcraturcs. Spnrtion took from 5 to 20 mins.,

depending on the composition of the acid used.

None of the nitrations unCcrt ,kcn Pith the sui-tchnicnl

plant fumed-off, or needed to be droyncd for fLCr of I fume-off.

/Thc-6-

The ester was transferred after one pre-wash to the leadwashing tank where it received three alkali and three waterwashes each of about two parts of wash liquor to one of ester.If on completion of these washes the heat test was not above10 minutes, one further alkaline wash and two water washes wereadministered. At first sodium carbonate, later sodiun bi-carbonate was used for the alkali washes, and in the lastnitrations sodium sulphite as well. It was found that a 2 percent sodium sulphite/sodiuia carbonate wash had a most valuablestabilising effect, and cnablcd the washing to b.. reduccd totwo alkali and two water washes.

If any acid fumcs werc present during filtrotion, theDLGN absorbed enough of them for its heat test to fall.

Heat tests of up to 20 minutes were obt,-intd, but wcregenerally 10-13 minutcs.

A selection from the dattd on thsec nitri:tions is given inTable 2 below. The figures givLn for the coripositions of thewastc acid etc., ar( obtained by an,)lysis; soi.i of the DEGN inthL waste rcid therefore appeors as nitric acid .nd "'v;ter;", ascxpliined elsewhere in this rLport.

It will bc sc,n thnt the highest yield recorded is 87.8,.)*The lowness of the yields is dut to thL 1,rge numbcr of bulkywashes needed before it w's found that sodium sulphitc solutionseffectively stobilised the productw.

The results in Tnble 2 bear out the deductions mcde fromdetcrminotions of the solubility of DEGN in w!stL ncid, whichw(rf discussed in Part II of the rcport.

The results set out in Tnblc 3a nnd 3b rLfcr to experimentalnitrntions corricd out at a loter date, when information concern-ing Gcrmon methods of manufacturing DEGN became avnilable.

At Krurmel a batch process was operated in which DIG wasnitrated with 2.90 parts of a h\:Lixed Acid containing 65,, HN\0 3 and35,, H2S04; the continuous process operated at Bolitz used3.18 parts of an acid of coh,position 6Z> HM 3 , 35,, H2SO4 and$ H20. The yields reported were higher than those calculatedfrom these figures, with the use of the graphs and data givca inPart II of this report, and so a fresh series of nitrationswcreundertakLn with the 20 lb. nitrator separator, to check thisinformation. At thL same time similar nitrations were fiade with2.57 parts of mixed acid of composition 7a, HN03, 2Q-1 HN03, and2.70 parts of acid of coiposition 69.0' IMO3, 29.5 H2S04, 1.5'j'H2 0, thsec being those which give a waste acid of 26.5:. watercontent and a ncar-minimm solubility for DEGN, as found in PortII.

The corpositions of the waste aci&s and acid esters producedwere obtained by analysis and were also calculated from the avail-able solubility data, ind the results compared with the str)tedcompositions.

The life of the waste acids before they fumed-off, was alsoinvcstigoted; these results arL reported in the appropriate sectionof Part II.

/The-- -

The results show that the yields obtained agree reason-

ably well with those calculated, except in the case of the

65/35 acid. The yields by the German techniques were not as

high as claimed, being slightly lower than those using our

nitration acids; moreover, larger vessels would be required

for the same output, owing to the higher acid ratios. On the

other hand the separations were cleaner than ours, and the

waste acids w(re aorc stDble and safer to handle. This latter

is no advantage if the waste acids are denitratcd immcdiatcly

after separation, as could bc the cDse in a continuous process.

/Table 2.

4-

0 44.

HEH

0 1 1 0 .3

~o

H0 WN-

8 r-I r-I r0 CN 00 %00 0 C

H0 H

HCH

0

0 0 oD o

0 0 Ul%LC L* f* L*

(F)c4-OH

<* 0

P

0.2-D

Cd 0 C 30000

E . 1("1 C\lC\H1 HHHHHH

H'~4.r'i~or d A.nH

IDIg 0 0)

o ~ ~ ~ ~ o o -oCSA'COC)

'HO H H

H4C H H Hi.E-i

Pt

0 Cd C 0-

r- a)c) N .

0 0 l

(TN O H r-0 .

~t) r-z C \l H C\l~

4) o3 a c ko'.

oo nd

o-, 0 1 N0 0-

r-:c 4Ci H. 6 & r2riH r491'\\ N

Ci m_ _ _ _ _ __ _ _ _ _

TABLE 3a.

DETER1INATION OF YIELDS FRO. -YERiAT NITRATION PROCEDURES.

EXPERILM TAL DATA.

Nitration . . IV.

JLG used 13 lbs. 13 lbs. 12 lbs. 11 lbs.EVA used 33 lb. 0 oz. 34 lb.14 oz. 34 lb.13 oz. 34 lb.14 oz.

F.J.A.pro-duced. 56 lb.2' oz. 56 lb.5' oz. 57 lb.71 oz. -,C lb. oz.

Disp.Acidused. 46 lb.071 oz. 45 lb.8 oz. 44 lb.2 oz. 42 lb.21 oz.

W/A bydifference 10 lb.2 oz. 10 lb.14 oz. 13 lb.5'oz. 14 lb.21 oz.Add '/A run

off 3 lb.9 oz. 3 lb.0 oz. 3 lb.10 oz. 3 lb.llo.

Total W/A. 13 lb.lloz. 13 lb.14'oz' 16 lb.15'oz. 17 lb.14oz.

DEGN after PA,lproduc ed.Weighed 22 lb.4 oz. 21 lb.13 oz. 20 1b.41 oz. 13 lb.6 oz.From baffles 11oz. 5'oz. 2 - oz. 6 oz.samples 1 oz. 1 oz. 1 oz. 1 oz.Total. 22 lb.61 oz. 22 lb. 3ioz. 20 lb.8 oz. 18 lb.13 jz.

Composit ion.

DEGN 99.4k 97.35,, 99.65i 99.15,HN03 0.0 .30 0.30> 0.0', 0.31.,H20 0.5 2.35',, 0.G5. 54,

NitrationConditions.

Brine InletTemp. -160C. -130C. -80C. -160 C.Initial Temp. 40C. 5°C. 40 C. 40 C.Maximum Temp. 15.5 0 C. 15.5oC. 15.80C. 150 C.Final Temp. 8.50C. 60 C. 3.50 C. 80C.Time of Nitn. 321 mins 32" rains. 32- fains. ,,.71 ),ins.Time of Scpn. 10 -iins. 10 iiins. 7 i,in. 10 i,ins.

Remarks;--

Separation No fuming or SomG fuming, Very clean Scpn.good,marked temp. no markcd separation, no fuming

rise. tcrap. rise. no fuming or undueor tc-ap.rise. tf-mp.risc.

After Sepn. Sonic after Somc after No after- Very litt1csepn. in scpn.(more separation aftcr scp-wastc acid. than in I) noticed. @ration.

/Toblc 3b.• - i13

Mt1TMM,n TION OF YIELDS FRUI GE

qmTmay of ati ties ana cgnpositions

R 4 ... ... .

l\1 ix9 ars 9 Cd Acid V as t N cid

calcd. 72 28 - 0 2.57 54.4 13.5 5.3 26.8

Found 72.25 * 28.0 12 0.57 2.52 NOT0.12 -. 57 252 ]~ TA'1LYSED

Calcd. 69.0 29.5 - 1.5 2.70 54.35 13.45 5.3 26.9

Found 69.0 29.5 0.65 0.85 2.70 NOT ZT,YSED

Stated 65 35 0 2.90 60 8 6-7 19-20

Calcd. 65 35 0 2.90 61.05 10.65 - 6.7 21.6

Found 65.8 5. 0.15 -1.00 2.90 61.45 10.9 0.30 5.3 22.05

6c. - 1 2 - -- -2-Stated 62 L 5 - 5.18

60.1 12.8 - 4.5 22.6Iv ICalc. 62 35 - , 5.18 56.9 12.75 - 5.5 24.8

Found 61.9 34.45 0.15 35 3.18 585 12.7 0.77 4.5 23.53

m Stated yield of finished DEGN after completion of washing.

-1a

I

LDS FRUAi GEMIX NITR TION PR0DURES

bo-npositions, with Cao.- .pariso n of Not Yields

-,cid Ester II Yield of DMlN Not yield DEGN/ 7--- / f after separation after two prewashes

. H0 _. .& p ar_ts . . theor.

3 26.8 1.3 21.3 77.1 0.3 178.2 173.3 93.6

NCT T,LYSEDi 172.4 93.2

.3 26.9 1.3 21.1 77.5 0.1 177.3 172.8 93.4

NOT TAhLYSED 171.C 92.4

-7 19-20 - - - 170. m 91.9 3

17 21.6 2.3 22.7 75. C C. C 174.C 168.8 91.3

o3 22.C5 1.2 20.75 I 77.1 0.95 17c.9 92.35

5 22.6 1.8 23.3 73.6 1.3 176.9 , 169.5 ,, 91.6

5 24.8 1.5 22.2 " 71.55 4.75 174.6 169.5

.5 23.53 0.95 19.1 77.4 2.55 171.1 92.4

.ng.

-1I

11. 125 lb. Batch Pilot-Plant.

1. Introduction.

125 lb. Batch Plant at "Voolwich.

.In order to cope more easily with increasing Departmentalrequirements for DEGN a 125 lb. Batch Plant was constructed andoperated at Woolwich from November 1943 until work of thisnature was transferred to A.R.D). Waltham Abbey in Sep,:t. 1945.During this period approximna;ely 7000 lbs. of D-,GN were produced.The plant comprised a stainless steel Nitrator-Separator of 20gallons copncity fitted with a leod cooling coil, stainless steelstirrer and bottom run-off vin a sight glnss und cock. Thevessel wos virtually identical to.tha-t now in use at 1:1Jalthnm Abbeyand la:ter described in detail. It is only proposed, therefore,to refer briefly to the following s-peci,,l points in conncctionwith the opercition of tht- pl,-nt.

(i)' The quantities of u~tr:snorm--,lly L,,,Yj)oyCd vlcr,. 75 lb.of D~LG 1nd~ 200 lb. of L'.ixc6 hciC (1-NO34~i 2 SO4 . 72/2413).

(ii) Cooling wa-s k_ffccted by pa,-ping icc& W tLr through the.l*d coil, -ind the ice- us"get_ i,.ountLd to 5-7 cvwt. -,e(r ch rg(

(iii) The CiL w>'S CO01C6 initi-211y to -bout 100C. nC the D!iGrun in ovcr -I)eriod of -,bout 35 ,iins5. The t,-,i-,,)r turc of th.reriction mixta~rL USU'-11, rose to 15-1800. n ii 's fin lly cool(-ddown to .-bout 100C. A -,D(rio(" of 10 iiis. vv is 11OxWCA foreLp-r-ition zind then th. loVVLr 1layer of S-,ont Acid w,-s run offrind Co1lCted in sta-inless StL l wiCkLtS. No ttciapt w-s i,deto re.COVer the nitric or~ sulphuric- .cids froiIi thL SI)Lnt ',cic a ndon -iccount of its poor st ibility it w,,s ifrdldi -t(..ly dc-stroy(d bythe process of "IfuL.ing-off" of three SUCCLssiV,_ portions in a-niron pon (20 grillons c-L)ecity) out in the_ ol)n. The dccoripositijnwase st.,rtcd by injccting liv(_ StLe. into the s-Ient tcid ;-nd th(time t,,kLrn to furaL-off vils 5-10 mine. The rcil-ining rcidresidUeS NVerL dcstroycd in a Lii,,c Pit.

(iv) Thc Acid ESter rcf.aining in the nit rot or-ELp,1-r,"tor w.-srun into buckets (,nd poured into i50 ga,ll. st.linless SteL ta-nkfilled with waitcr ind icc (opproxirivtcly 5o/5o). The contL.ntsof the tank were ngitatcd with comprcssed 7~ir for 10-20 iains.[Ifter which timc the tc;iipc.roture r;lngcC froiai 10-150C. The _ '.GCNwios a*llowcd to scpa-rote, the wtcr siphonLd off, and thL Sep-ri7tLd

* oil tr,ansferrLd to a. Wiooden Jirshinf Va t. Norrally, thL DEGNfrom four nitrations (i.c. onL d,-y s m-nuf,-cturL) wi!-i ncuaula-ted.-nd ifter a standnrdised trea,ti:eint of six voshLf: " He:-)t Tca~t inc xcces of 10 mine. vvas obtained. DeLtailS of the v4-rious wvishingproCLdurcs which hive beeLn eiaploycd aire giVen in Section 4.The wanshcd, et,ibilised ester vic's filtLre-d through p, Per to reLioVeextraineous raoisturc, a- nu -,tored in c rboy,,_.

(v) A fin:l yield Of bout 1221 lb. of ~L*JGI," nc' obt in 6 fro-.-75 lb. of DLG corres-,onding to 3S'_- of theory. Poorer yieldowere obtained if iaanipulation at the ti,-e of running off theSpent Acid were faulty, e.g. if run off at too fast -6 rate so-.Aemter was inevitably carried away in the vortex.

13~ lb Batch

125 lb. Batch Plant at Ialtha. .,bbe..

This plant was installed in the old F.G. Section (No.1 Hill)

at 1.altham Abbey during the latter part of 1945 anc caue into

operation during 1946. The nitration unit is based on that

formerly used at -'oolwich jiad where possible, the existing 11.G.

services (e.g. acid lines, buildings, gutter coim,iunic,)tions ctc.)

have been utiliscd. The two original iE.G. nitrr-tor-sa, r-tors,

the pre-wash t,nk n the baffle tank, , rcuiovuL fro,, the

Hill to pcrwit the inst,llrtion of thc DiLGF unit. Th buil6ings

brought into usc v-r-"

Solution House (for sodiuia c,rbon -t. -n6 sulihit.solutions).

iAcid Fixing Hous(.Chorgc house.Nitrating HousL.Igg 1ouse (for SLj)nt Acid).

I ashing hous.Pouring-On house (not now used).Viash XTt .r Settling House.Acid Dtnitration House.Ivia gr zinc.

A detailed dea-cription of the plant is giv(n below. A Flow

Sheet, showing the cycle of opcrotions -ad thL qu-ntities of

DlPtericls involved, is reproduccd in Drfving No.1 (-,.R.D.8154).The following drnwings of the pl-nt rc repro uccd:

Drnwing No. Original Rcfcrence DLtnil Dr:-1wing. .. ......... o f . -

2 ED(O)1868/I/5 C General nrr,ngcmcnt of plant.

3 /21 Layout of piping inNitration House.

4 /15 iixcd Acid Gauge Tank.

5 /14 DLG Gauge Tnk.

6 / 1 6 A Arrvngt_ii,nts nd dtailsof Nitr,'tor Drm ning Volve.

7 / 2 0 A hrr--ngcints ind det-ilsof Nitr: tar Cover.

8 - Prcvmsh T7nk.

9 -,Vastc Acid T.rnk.

$uvply and Handling of Raw I:tcrials.

DLG. This is kept in the old Glycerin Storc, -nd on, drunat , TjiuI( UlCv:atcd to the Ch,rgf -housc. The Ch7rgc House supply

tank is n vertical tank, c pacity 60 gllons, with dished cnds,

i.e. totally enclosed. The Glycerine Supply T.nk, r rctangula"

galvnised tcnk with loose fitting covtr, wu,s found unsuitable

owing to the extrCm hygroscopicity of DIJG, and has becn removed.

ACIDS" 98-99, Nitric Acid from the B 'mag Towers, R.D.X. Acid

Factory, and 20;, (Trade) Olcum, art mixed in a smnll pl,int spcciclly

/installcd

- 14 -

installed in the Acid Y'ixing House. This consists flead and mildsteel gauge tanks for Nitric and Oleum , (original ITG iYiixingplant), froii which the acids flow throu7h feed boxes, contain-ing calibrated glass si)hon tubes, into a sm;all vertical leacdruia, which the two streaiiis enter tangentially. Froi-a thebottoi. of the iiixing vessel, the acid passes to a -verticallead cooler, capacity about 40 gallons, fitted with lead cool-ing coils (circulating cold water), air agitation coil, andfume off-take. The cooler has a bottom outlet and also anoverflow, from either of which the rixed acid may be pumped toa 1000 gallon storage tank. This arrangement permits eitherlarge or small quantities of acid to be mixed. From thestorage tank the acid is puraped to the Hill, and is receivedat the Charge House in a 130 gallon m.s. tank, whence it ispumped through a metafilter into either of two 50 gallon verti-cal stainless steel supply tanks.

1iASH SOYUTIONS: Sodium Carbonate solution and soft mainswoter (softened by boiling) are. prepared in tanks at the N.G.Solution House, filtered (through flannel), blown from m.s.eggs to the Charge House, and stored in golonised tanks of800 gallons and 550 gallons capacity respectively. These canbe heated by steam coils. Supplies are piped to the NitratingHouse and Washing House by gravity flow, controlled by brassplug-cocks.

Untreted river watur, used for the prcvwshks, is puipedto ai water tower ond piped by.grvity to the Nitrating House(all origin-l N.G. plant).

OTHLR SERVICES: A D)iesel comprLssor sup lies coi-prcesscdair at 30 lbs./sc.in. to - ring i&-in in the Nitr:ting House, tothe ,.shing House, ind to othLr buildings whLre r,quircd.

Cooling brine for nitritioll etc., is obt ind by p-ssing30-351,, Calciui7 Nitrate brine (containing 71,- Sodiuma Nitrate)through an aminonia refrigerator. A storage tank on the hill,capacity 2,560 gallons to thf overflow (continuous circulation),provides a tead of 20 ft. at a temperature of -16oC. or below,

to the nitrator, pre-wash, and spent ocid tank cooling coils,although an inlet tcmpfrz,ture of -80C. is sufficient for safeworking. In view of tht fact thOt sowe of the cooling coilsare stainless steel, it was morc dcsirnble to use n CalciumNitrate brine than Calcium Chloride brinc.

" The Nitrating "Hill".

This comprises the N,G. Nitrating House, a roughly circularbuilding,. wooden construction, tr,verscd and mounded on allsides to a height of 40-50 ft., the Charge House ond BrineSupply Tank situated on top of the mound, and the Spent AcidEgg House, outside the mound on ground lcvll.

The CHARGE HOUSE, a wooden building partly supported on picrs,is divided into two rooms, one for acid storage, the other forDEG and wash solutions.

The acid room contains one 150 grllon receiving tpnk, pump

and mctafiltcr, supply tanks, and the Lixed Acid G,luge Tnnk

(Drawing No.4). All the vessels and associated cocks and piping,with the exception of the receiving t,nk, -re of stainless steel.The sup)ly tanks can be fillcd and ehl-ptiLd indcpendcntly;

/their15

their bottom outlets are connected by a comiaon line to the

Gauge Tank, which is controlled by stainless steel (s.S.) plug

cocks on inlet and outlet. A 1 in. s.s. line leads from the

Gauge Tank to the Nitrator, and a second 1 in. s.s. line from

the supply tanks, passes direct to the Spent Acid Tank. The

Gauge Tank was originally calibrated 190-210 lbs. and 222 -242 lbs. of mixed acid (at 200C), i.e. two sccles, the upper 32

lbs. greater than the lower; by this means, with one filling

of the Gauge Tank 32 lbs. of mixed acid were masured and scnt

to the Spent Acid Tank in the Nitr:ting House for fortificationof the waste acid produced during nitration, and then the mainbulk of nitration acid sent to tho Nitrntor (Y.B. The 32 lbs.

of acid for the Spent Acid Tank was run through the Nitrator).

The solution room contains the original N.G. soda solutionand soft water tanks, vnd the (new) DEG t-nk. The DLG supplytank has n bottom run-off controlled by a 11" s.s. plug cock,connected by a lead line to the DEG Gvugc Trnk (Drawing No.5)which is mounted on a shelf above the nitrition plntforn in the

Nitrating House. The GugL Tank is c libr-tLd from 70-80 lbs.of DLG at 200C. The inlet and outlet irc controlled by plugcocks, with a drain cock on the inlet side.

The NITRATING HOUSE, has -. wooden platform along one side,opposite the entrance, which constitutes the working level for

nitration. An escape door at one end of this platform leads tothe top of the mound. In front of the plntform a steel frame-work is erected, which carries the DEGN V<itrotor. The Prc-wnshTank stands on a lower platform in front of the nitrntor, and thcSpent Acid Tank rests on o stool on the floor. TWO bofflc tanksconnected in series, also mounted on wooden stiols, overflow intoo gutter leading to the Wash Water Settling House. Steps lod

from the floor to the Pre-wash Tank platform, on which there is

working space, for separation and pro-washing, and thence to thetop platform.

Underneath and accessible from the main floor, is a well,which contains o large elliptical drownirg vat, (the original N.G.Drowning Tnk), containing 3,500 gallons of water.

The NITRATOR is a 20 gllon s.s. vessel (foroorly a SodiunAzide Vessel), with a dished bottom, and a top flongL, supportedin the framework by two side lugs. The contre of the dishedbottom has bLen cut out, ond n n w s.e. plto weldcd in (DvQaingNo.6); this has a 1 in. outlct, off ccntre, for normal runningoff, and a 2 in. control outlet which is normnlly closed by r flapand scaled by a pad of "Nitresto"; the flp volvc c-n be rlecscdby a lever mechanism, to drown the charge.

A flat s.s. cover (Drawing No.7) is bolted to the top flange.The vLssel contains a two-bpnk lend cooling coil, total colingsurface 20 s'q. ft., and a centrally mnuntod s.s. stirrer shaft,which carries two double bladed impellors, one near the bottom,the other adjustable in height. The stirrer shaft is supported

from the superstructure, and is driven at 200 r.p.m. through aright-angle drive from a H.P. motor housed in a casing on theoutside wall of the house. The brin, inlet ond outlet, the MixedAcid line, fume off-tnkc, and DEG feed line all pass thrqugh thetop cover, which also supports a thermometer pocket (totally

enclosed, oil filled, containing a mcrcury-in-glass thermometer),and has a 6 ins. diameter inspection window. There is also

/provision

- 16 -

provision for the insertion of the bulb of a recordingthermometer (mercury-in-steel), which is housed outside thebuilding. The DEG supply from the Gauge Tank is controlledduring nitration by n I in. plug cock and show glass, mountedabove the nitrator cover. The other controls for nitrationcomprise gunmetnl plug cocks and thermometers on both brineinlet and outlet; the stirrer motor is controlled from aswitch just outside the escape door. The drowning mechanismis operated by either of two hand levers, from the top platformand from the pro-wash platform, and also by n line and pulleysystem from outside the tunnel entrance. The drowning outletdischarges through a funnel and pipe into the drowning tank.

An air poker is kept handy for emergency agitation in casethe stirrer motor should fail.

The 1 in. bottom run-off has a sight glass illuminatedfrom behind by n 15 watt lamp in a flameproof cnse, and a threeway s.s. plug cock, by which separation is observed and con-trolled. The cock is arrnnged so that the OFF position is inbetween the two OPLN positions (a) to Spent Acid Tank, (b) toPre-wnsh Tank. When the nitrator contents have separated, thethree way cock is first opened to run the lower acid layer tothe Spent Acid Tank, ond then, as the separation level bocomcsvisible in the sight glass, revcrsed, so as to run the esterlayer to the Pro-wnsh Tank. This cock, which is n vital part

of the pldnt, is constructed to have no internal "pockets' whichwuld retoin unstable Gxplosivec.

The SPENT ACID TANK originally installed for the DEGN plantwas much too large, and his been replaced by n smaller one,capacity about 20 gallons, dimensions 16 ins. diameter x 24 ins.deep. It is of lead, and hrs a sloping bottom and domed coverwith n ccntrnl fume off-take. The separated acid cnters throughthe tOp cnvLr, air agitation and brine cooling provided by nlend c"il of about 9 sq.f. surface area, being used if fortify-ing acid is present. A thermometer sleeve posses through thetop cover, which also has a glass inspection window. The acidis run to the EGG HOUSE through n 2 in. lend bottom run-off pipeand earthenware cock, and trucked from there to the dcAitrotionplnnt.

The PREWASH TANK is fabricated from s.s. sheet, w0dcd up,with a sloping bottom ond top flange, dimensions 2 ft. 4 ins.internal diameter, 2 ft. deep on longest side. The boltedin domed top cover has a central fume off-tnke, and cut-awayinspection hole, normally closed by a rubber apron, through whichpass the supply pipes for hard river water, sWft water, and sodannd sulphitc solutions, which are controlled by brass cocks orwater taps at the side of the tank. The cover also supports anoval s.s. cooling coil, set obliquely in the tank, of 10 sq.ft.cooling surface, and a therw,ometer sleeve. An air spray coil isfitted for agitation. The tank has a bottom outlet for runningoff the DLGN after washing, and a second, higher up, for runningoff the skiiiiaings. The skimmer consists of a P.V.C. funneland hanc,le attached to a pure rubber hose. The DEGN and watcrsare run off through rubber hose pipes which are tied up to durr,nyqfaucets when not in use. The faucet for the DEGN hose has anair pipe attached, for freeing the hose pipe of unstable explosiveduring washing. The run-off hose can be used to drown the chargeif necessary.

- 17-

,'fhen the DLGI'§ is alkaline (generally after one soda

wash) it is run fro.- the Prewash Tank to the '-JAShI!'G i0UE

down the N.G. GUTTER. This is of lead, double walled, length

176 feet, painted with a rubber base paint (as are also the

lead baffle tanks) to preserve the lead. The gutter warm-

water jacket is not used for DEGN.

Fumes from the nitrator and spent acid tank are with-

drawn by 2 in. lead piping through earthenware "folff's bottles

situated in the annular space between the Nitrating House wall

and the traverse, and through carthenware piping to packed

earthenware towers, where they are scrubbed by water circulated

by PohlE air-lifts. There is one tower to each fume run, and

they consist of four 2 ft. diameter sections. The suction is

maintained by an air-fed aluminium jet in the chimney of cach

tower. The Prowash fumes are withdrawn by a lead pipe through

a catchpot and out to the atmosphere, suction bcing maintained

by a Norki (silicon iron) air jet.

A cord and pulley system shows on an indicator in the

heuse the level of brine in the header tank on top of the mound.

This is non.,ally always full to the overflow.

ihIshing House.

The WAShING HOUSE is a circul,r wooden builCing, 20 ft.diameter, -tr,verscd -nd mounded over on -ll sides to - height

of 20-30 feet, with a tunnel cntrnncc. A platforfm on one side

of the house originally supported three N.G. wishin- t-nks.

One of these has now been replaced by a smller one of similar

pattern, suitable for washing 1 - 2 charges of DLGN; the

larger tanks crn be used for washing sevcr-l b tchcs together.

The lrgc tanks arc 3 ft. 9 ins. di-1etcr x 3 ft. 2- ins. deep on

the longest side (sloping bottom) and c;,p.city 209 g-llons.

They are of le:.ad, havc two run-off pipes, one -t the bottom, ond

the other higher up, to which 7 pure rubber skiai cr is attachcd,

and nrc covered over by - cloth cover except ,t one point where

a ftuae pipe is burnt on to the side of the t nk. An nir spray

coil is fitted for agitation. The a,.ll tank is similar, but

hos a diameter of only 24 ins. The skimmings p,ss through an

oval baffle t7nk st.,nding on a lower platform, and then down o

lad gutter to the WASH WATER SETTLING HOUSE. The fumes pcss

through lead catchpots and out to the atmosphere.

Originally it was intended that the washed DEGN should be

piassed to the POURING-ON HOUSE for filtration in the N.G. Filter

Tank, but it was found convenient to usL a smaller appnratus

which is now installed in the Washing House. This consists of

o lad tank standing on the baffle tink platform under the wash

tank run-off pipes, the filter standing in a lead cradle inside

the tank. The FILTER is a stainless steel funnel, 15 ins.

diameter with straight sides 10 ins. deep in which arc super-

imposed two filter pads, consisting of small sponges scwn up in-

side flannel covers, cut to shape and stitched on o sewing

machine. This will filtcu 4 - 5 charges (500 lb) of DEGN before

becoming saturatcd with moisture and DLGN, when the spongecs -r

removed, washed and squcczcd out, and replaced. From the t-nk,

which acts as - rcservoir, the filtered DLGN is measurcd out,

25 lbs. ;t - time, from n N.G., cordite p;ttcrn, le:ad burette

which st.inds on a lcnd block burnt to the lead floor, into 4

gallon l'cquLrcd pressed Etccl c ns. This c --ntity (5 lb..) is!

/considcrcd

- 18 -

considered suitable for present requirements having regardto safety and ease of handling, etc.

The wash liquors passing to the Wash Water Settling Houseflow into the Wash Water Settling Tank (original N.G. plnnt),of 5,600 gallons capacity. This is emptied periodically andany DEGN found is destroyed. Any DEGN found in the cotchpotsin the Nitrating House and the Washing House is also destroyed,but DEGN found in the baffle tanks is washed with subsequentcharges and recovered.

3. Operation of Plant.

At Woolwich, when iced water wns used for cooling, thenitration took about 35 minutes; nt Waltham with the ommoniirefrigeration plant, brine is circulated through the coils,entering at -100C. or below, and nitration takes only 20 mine.The acid (200 lbs.) is initially cooled to below 100 C., andDEG (75 lb.) run in a slow stream from the gauge tanks throughthe sight glass, on to the surface of the acid, being rapidlydispersed by the efficient stirrer. The temperature is nllowedto rise to not above 150C. as the nitration proceeds.

Separation occupies 10 minutes, no accelercnt bcing added.The lower layer of waste acid is then run into the waste acidreceiving tank until the level of seppration reaches the showglass. The three-way cock is then reversed and the acid esterrun into a measured quantity (7 gallons) of water contained inthe s.s. prewash tank, with air agitation and brine cooling tokeep the temperature below 200C.

The duration of the first wash is the tin:e of running infrom the nitrator, about 10 minutes being required, after whichthe first wash is skimmed off and set aside. A second prewashfollows, the water from which is discarded after skimming off.The DEGN then receives one alkaline wash in the prcwash tankand is then transferred to the N.G. pattern lead washing tanksin the Washing House, for final washing and stabilisation.

A small washing tank is used for a single 125 lb. batch;the larger, original N.G. pattern tanks, will acconmmodatc sixbatches of DEGN, these being then washed together as a single750 lb. batch.

After a satisfactory heat test has been reported, theproduct is filtered, measured out 25 lb. at a time, in the leadN.G. pattern burette, and stored in lacquered cans.

The cycle of operations from the commcncemcnt of nitrationuntil the DEGN is trinsferred to the Washing House, after afirst olkrlinc wash of 45 minutes duration, occupies 2 hours,so that three nitrations can be complotcd with csc in a normalworking day, nnd the plant is therefore cqpable of producing1500 lb. DLGN per week, allowing four days only for nitrating,and one for cleaning up etc.

The method of trcntmnt of the separqtcd waste acid,originally adopted, was to fortify it in the waste acid tankwith sufficient fresh mixed acid to reduce the wotcr content to20k,, and ensure safe handling before denitrntion. It wasfound, however, that the resulting composition was not very suit-able as n denitration tower feed acid, dnd various modifications

/of- 19 -

of this procedure have since been tried. For the resultsof these trials reference should be made to Section V ofthis report.

4. Comments on the Design of the Plant.

The general arrangement and design of the plant has, inthe main, been found quite satisfactory. Some minor modi-fications which have been made to individual items of plantare discussed below.

One alteration, which is required to enable the prewashliquor to be separated more sharply after the first prewashing,has not been made; this is because the work on the recoveryof these liquors from the batch plant, was discontinued whenthe continuous plant came into operation.

Nitrator: The mechanical agitation and brine cooling areadequate for the purpose of nitration; the only difficultyexperienced with the nitrator has been blockage of the run-offpipe leading to the thre-wny cock. This was duc to sludge inthe nitrating acid, picked up from lead acid piping. To elim-inatc this a Mctnfiltcr wcs instnlled in the Charge Housc andthe mixed acid filtured before use. As an odditionnl precautionthe 2" run-off pipe was replaced by a 1" bore pipe.

The lead cooling coil showed some signs of waor aftcr 70nitrations, but was not seriously impaired. A stainlcss steelcoil would probably have a longer life than lend, for thisservice.

Separation Cock; This is a vital part of the plant, whichhas received considerable attention. Originally two full-bores.s. plug cocks, lubricated with a thin smear of non-reactivemineral jelly, were used, one on either side of a welded s.s.tee piece, so that the run-offs to the waste acid and prewashtanks were controlled separately. Subsequently, a single three-

way full-bore s.s. cock, similarly lubricated, was substituted.With both arrangements, some scoring of the plugs occurred.The plug of the three-way cock was then plated with chromium,and relapped into the barrel, which was not treated; after 25nitrations, hardly any scoring has occurred between the two dis-similar surfaces. A three-way armoured stoneware plug cockwas also obtained, but this has not been given an extended trial.

Waste Acid Tank; The original waste acid tank was too large,and when fortifying acid was used, most of the cooling coil wasineffective. The tank was replvced by a smaller one of 20 gall.capacity.

Prewash Tank: The originAl design wqs modified by reducingthe height of the vessel and also by making the circul7r coolingcoil into an oval shape, thereby mYking it easier to skim down.As it is still difficult to obtin really sharp separation withthe skirmaer, which is necessary in separating the first prewashliquor, it is considcrod that o siphon pipe would bc more suit-able. The only alternative would be to run-off the lower DEGNlayer, leaving the acid behind, but since further prc-washing isrequired, this could not be done conveniently unless a secondp.w. vessel was introduced, and the levels of the precedingvesscls rvised correspondingly.

,/hIa shing-L20 -

Washing Tanks, Filters: The batch washing tanks andfilter used for DEGN are of a similar pattern to the corres-ponding N.G. plant and no alterations were necessary.

Fume Catch pots: Although DEGN is more volatile than N.G.the same fume sytem was used, and no alteration is thoughtnecessary.

5. Plant Investigational Work.

The chief interest in operating the pilot plant was todetermine the net yield of DEGN obtainable on this scale of

production. This depends on the nitration conditions, theahiount of washing necessary, and the mechanical and volatilitylosses, etc.

The nitration conditions odopted were those recommended inPart II. viz: 2.57 parts of 72/28 mixed acid; o serics ofnitrations was -lso undertaken, using 3.18 parts of 62/35/3acid, principally to gain experience of handling and reworkingthe waste acid produced, ns this process was envisaged for theinitial operation of the Schmid plrnt. For this purpose it wasnecessary to reduce tht qu.ntity of DEG nitrated, from 75 lb. to70 lb., bccause of the higher proportion of mixed acid. Inall, 15 nitrations were mnde with this acid, resulting in onaverage of 115 lb. net overall yield per charge (88.9 V). Asexpected, this was less than the yield obtainable by the originalprocess (with an equal nuuber of washes after sept:ration), butthe waste acid remained stable for a longer period, and was moresatisfactory as a denitration acid.

Washing: The scheme of washing originally used, consistedof a single water prewnsh with about 3 parts water to 1 partDEGN at 10-15oC. for 10-20 minutes, followed by scveral alkalincwashes and at least two final washes with soft water, at roomtemperature. The introduction of alkaline sulphitc enabled thenumber of alkaline washes to be reduced to a minimum of two orthree.

It w s necessary to reduce the amount of prewash water used,so ns to obtain the acid wash liquors at a recoverable strength,

(30-35,cl HN03), and under these conditions a second wotr wash isprobably desirable before washing with alkali.

Experiments were made on the plant, using one and two pre-

washes, in a series of batch nitrations; sarples of the DEGN andwash liquor were taken at intervals during each wash and analysed.The results, Table 4, show that while a fairly long time is re-quired to reach the equilibrium distribution of IHN03 betweenthe DL,GN and water phases, it is unnecessary to prolong each wash,

since two short prewashes, each with a reduced quantity of water,achieve a reduction in acidity coraparablL with that resultingfrom one long prewash taken to equilibiium.

It was- reported that DEGN made in Germany, was given two

prewashGs, the first with a reduced quantity of water, to produce

recoverable acid; these were followed by one, or possibly two,

alkaline washes, with 2-3 parts of 5c") sodium carbonate solution

at 60oC., for only 10 minutes, and finally a cold water wash.This brief washing scheme was said to produce material of 20 mins.

heat test, and frequently 40 minutes./Ta2 4

- 21 -

0CadPARISON OF SIGL JM 71-OIAISH,2 l. ~

IV

Batch I II III Ist Prewash 2nd Prevash(45 - -a RsT (30 39ns)

Ratio of water to 1.725/1 1.41/1 -1.725/1 0.86/1 1.725/1acid ester

Acidity after - 0.43% 1.14;" 0.82/ 0.03i,10 mins.

Acidity after25 mins. - 0.351' 0.39%

A,cidity after60 mins. 0.31% 0.o8 0.24/

Final acidity ofWash Water (01) 11.4c 12.52% 11.8/,, 18.9/,) 2.14

Equivalent equili- below

brium concentration 0.06% 0.07% 0.07% 0.22% 0.01

of HN03 in DEGN(C2)

Batch VI VII VIII

lst prevash

Ratio wash water to 0.45/1 0.45/1 0.45/1 0.45/1

DEGN

Time of washing 15 mins. 15 mins. 15 mins. 15 mins.

Final acidity of 1.80/ 1.05% 1.95% 1.47/

DEGN

IZHT03 in Wash Water 24.1 28. 0 34.0%, 29.6%

(01)

Equiv.Equilibriumconcentration of Th03 0.54% 0.88% 1.80/,% 1.06%

in DEGN (02)

2nd Prewash

Ratio wash vater/DEGN 0.6/1 0.6/1 0.6/1 0.6/1

Time of washing 15 mins. 15 ains. 15 mins. 15 mins.

Pinal acidity of DEGN 0.34% 0.13/ U.2 8 % 0.10%,

Final acidity if 0.45 +o,6 = 1.05/1 partssolution used in a 17.6% 21.4% 20.2> 20.2/

single prevash, taken toequilibrium.

Bquiv. equilibirum 0.18/) 0.35zn 0.28 0. 28oncentration of HN03

2DEGN

-22-

If this was indeed the case, the design of the washingplant must have been extremely efficient. We have triedshort hot alkali washes, and although the high temperaturesused undoubtedly assisted emulsification, it was found thata longer period than 10 minutes per wash, was certainly re-quired under batch conditions.

The number of alkaline washes required during the earliestruns with the plant, was sometimes as many as eight or ten,this being partly due to the newness of the plant and the needto establish the optimum proportion of wash solution to DEGN.After a few nitrations it was found possible to standardisc thewashing conditions; 3"-4% sodium carbonate was used for thealkaline washes and the time of washing was 45-60 mins. Theaddition of sulphite solution to give 2;4 sodium sulphite in thewash was found to reduce the number of washes required. Thefinal washing was with softened water, as normal N.G. practise.

The following sche(me was finally adopted as being thc mini-mum washing necessary to give a satisfactory heat test :-

(1) 1st water prewash with approximately 0.5 parts w;tcrper part acid DEGN, ollowing a temperaturc of up to, but notcxccuding, 300C. for ten iainutces.

(2) 2nd water prewash with 1-2 parts of water per part DEGN,warm, for 10-15 minutes.

(3) 1st alkaline wash with 1-2 parts of 31-4, sodium carbon-ate, warm, for minimum of 45 minutes.

(4) 2nd alkaline wash with 1-2 parts of 3 -47' sodium carbon-ate with addition of aj sodium sulphite, warm, 45 minutesminimum.

(5) 1Warm soft water 1.1 parts per part DEGN, 30-60 minutes.

(6) Cold soft water 1.1 parts per part DEGN, 30-60 minutes.

6. Yields and Flowsheets.

The best yield of DEGN which can be obtained under the mostfovourable conditions, is necessarily less than that of N.G.,because of the greater solubility losses of DEGN at all stagesof the process; in the waste acid, the prewash liquors, andthe alkaline wash waters. For this reason, while it is usualto obtain 93.7% theoretical yield in the manufacture of N.G.,the best overall yield of DEGN will not be greater than 91.5;'.This figure is derived as follows :-

(1) From the graphs shown in thL section on the solubilityof DEGN in waste acids, a nitration ratio of 2.57 p.-rts of 72/28mixed acid to one p-rt of DLG wrs selected as giving a waste acidin which the loss of DEGN is ot, or near, its mini,umn. Theoctuil loss is .063 parts DEGN pcr port DEG, or 3.7(' of thetheoretical yield.

(2) Assuming, as is "pproximat_ly true, th't the loss ofDEGN in prewashing is a function of the total nitric acid in theester lyer, then the ester corresponding to the above conditionswill lose .050 ports DEGN per part DEG, to the prcesh liquors,

/or- 2 -

or 2.7j, of the theoretical yield.

(3) In washing, the best that has been achieved on the

pilot plant is one further prewash, one alkali wash, onealkali-sulphite wash, and two water washes. Using one part

wash solution per part DEGN, and taking the solubility ofDEGN in the wash waters to be 0.4', this means a loss of approx.2.1V of the theoretical yield.

The yield of DEGN after prewashing, but before the alkaline

washes, will thus be 93.6% and the final yield will be 91.,%.These figures neglect losses in nitration due to the formationof byproducts, and also mechanical and volatility losses duringwashing and skimming. The best overall yield of DEGN obtainedin the laboratory was 91.8b while the figures in Table 3(b)show a yield after prewashing, of 93.21 on the semi-technicalscale.

On the pilot plant, only thu final yields have been assessedaccurately, it bcing more difficult to weigh the yield afterprewa shing.

The plant originally opcrated at WJoolwich showed overallyields between 87 and 8Q , but these were mostly obtained beforethe value of sulphitc washing was realised. At Waltham, withthe standardised technique described, a net overall yield of90.j has been obtained. This indicates that the miscellaneouslosses, including nitration side-reactions, volatility and mechan-ical losses, amount to 1.3%.

The flow sheets, Tables 5(a) and 5(b) are based on the

above figures; some comparable information relating to the batchmonufncture of N.G., is given in Tables 6 and 7.

/Table 5(n)

- 24

TMBLE 5

FLO T FOR T N,U,CTU OF D.

IZG Used M.-. Usage W.,. Produced Lcid Ester ist Prewash!Produced liquidsproduced

.257 pts. 126 pts. 231 pts. 1Q.i pts.. .p pts. ' pts. / pts. 7' pts.

Recovercd as DEGN 90.2 1H2S04 28.0 72 H2sO4 54.6 68.8 H2SO4 1.5 3.2 2.4 3.2

Lcst in 3.7 603 72.0 185 HN3O 13.5 17.0 ITO3 21.2 49.0 55.0 45.6

Lcst in rowash 2.7 DEGN 5.4 6.8 DEGN 77.0 178.1 3.9 5.:liquor

Lost in later washes 2.1 H2 - 26.5 33.4 H2 0 0.3 0.7 58.7 7.3

Hechanical etc.losses. 1.3. Wash water

used75. 6

pts.

- 25 -

Ti. BLE .5(

TUE OF DEGN BY THE B-TCH FRC(Sj B-SIS)

Prewashi Ester after 2nd Prewash Ester aftex Wash waters DEGN ProducedLuids 1st liquids 2nd)duced Prewash Produced Prewash

,. pts. 17b.5 pts. 177.4 pts. 172.4 pts. 700 pts, 17.p0t.pts. i pts. > pts.

3.2 - - - Soda wash 175 pts, From, Prewashing 172.4 pts.

45.6 2.1 3.4 2.0 3.4 DEGN 172.4pts. Soda/sulphite 175 Losses in

final 2.8 pts.washing

I 5,r 57.9 173.1 0.4 0.7 Soft water 175 Transport

losses 0.4 pts.

7 I7.3 57.6 175 Soft water 175 J l,echanical &volatility 2,2 pts.losses

h water Wash water Soda ash 14d used Sodium 3 " Net yield 167.0 pts..6 pts. 175 pts. sulphitc

- 25 -

FL0W7 SIEET FOR TIM MANUFACTURE OF DEGN I

DEGN Used M..iL. Usage W.A. Produced I Acid Ester Ist Prewash Esproduced liquors

59.9 pts. ___- _-_r_dee

59.9 pts.- 1 153.9 Pts. 7 .5 pts. 138.3 pts. 78.0 pts. 1opt. pts. pt.pts

Recovered as DEGY, 90.2 IH2SO4 28 43.1' H2SO4 54.6 41.2 1.5 1.9 2., 1.9

Lost in 3.7 IHN03 72 10.8 BNO 13.5 10.2 21.2 29.4" 35.0 27.3 z

3

Lcst in provash 2.7 M N 5.4 4.1 77.0 106.6 1 3.9 3.0 97.LiquorLost in later 2.1 H20 2 .5 20.0 0.3 o. 4 1 58.7 45.8

vw.shes

cchanical etc. 1.3 I Wash water

losses . used

100. 45.4 pts.

- 26 -

TL 5 (b)

ME OF DEGN BY THE BATCH PROCESS (gqj l SjS.

rewash Ester after 2nd Prewash Ester aftcr Wash DEN ProducedWS 1st liquors 2nd Watersced Prewash produced Prawash

pts. 105.7 pts. 107.1 Pts. 103.2 pts. 418.4 pts. 100 pts.pa. Pts.J P ps.

1.9 Soda wash 104.6 pts. From 103.2 pts.

Prewashing7.3 2.1 2.1 2.0 2.1 DEGN 103.2 Soda/sulphito 104.6 pts. Losses in

pts. wash final 1.7Soft Vlater 104.6 pts. washingSoft water 104.6 pts.

3.0 97.9 103.6 0.4 0.4 Transport 0.2losses

5.8 - - 97.6 104.6 Soda ash 8.4 pts. Mechanical 1.3andvolatilitylosses

ater Wash water Sodium sulphite 2.1 pts. Not yield 100 pts.used

ts. 1046 pts. = 90.2/s

- 26 -

TABLL 6.

Nitration of Glycerine (Waltham).

1500 lb. Glycerine used = 1 part

8250 lb. 1L.A. it = 5.50 parts.

Composition of-L.A. = H2S04 55.57 per cent,11N03, 43.44 percent, water 19::jpe_aant.

Mass of HN03 taken for nitration = 3580 lb. to 3630 lb.

U If " theoretically necessary = 3080 lb.

Excess = 500 lb. to 550 lb.

= 16.2 per cent. to17.9 per cent.

Waste acid :- 5950 lb. of composition H2S04 72.9 per cent.,HN0g 9.1 per cent., HN02 0.22 per cent., N.G.3.3 per cent.

Mass HN0 3 in W.A. = 541 lb.

" N.G. in W.A. = 199 lb.

= 4.95 per cent theory yield.

Yield = 231 per cent on glyeerine = 93.7 per cent.theory = 3465 lb. Theoretical yield = 4020 lb.

Prewash. (1) 1.4 parts water.

(2) 0.62 "

(3) 0.5 " soda 3 ' per cent.

Wash House. (1) 0.45 parts soda 3 per cent.

(2) 0.62 water.

(3) 0.62 "I

(4) 0.62 "

Total washings = 4.8 parts by weight.

HNO 3 lost in Pre-wash.

Mass of 1st. pre-wash = 4850 lb.

Concn. of HN03 in pre-wash = 3 per cent.

Loss = 145 lb.

- 27 -

P-1-oo ,0

0 Cj

C\ %"0 co0 r4u-H

oo CO O co C~

0~~~~ cor1::P0 4 ci 41 4-

LC4o '

co

,L-C 4x rdI 0,

w0 0 0

C\ r- N 0. A •0•d g r-I H

_ _ _ _ __ _ _ __.......

4*4JEll ~~ ~ ~ 4 4. d P1 N H r- !c04

co cli ON LA c <r-I r-!

* C\j. ' . D o o ,-*. O CCf c5-) O N C 0

-t _H 0-0 HP4j 0;i -\

. . . .i r- . . . . .. .

Cd CN r-I

cd P4 H PA -N N- \ H- 0

rd H Lrl Hn H. c

C\jJ

F4 00

qjr 4 -:

j.; rold o. 00 0

ld

hZi

LUKg

um I

-codc

hi...

S 70

im W.w

.. ..... . J .. ..-

Lj n!

LC t 6p.*~ dr

IaI

zi FiFN

~~4J

z~~~ E!11f

-' ----- n

_

fit'TCMM

®G®®®G® ®G®@®®®®©Of

SO

V-.

.. .........

"is -

I rkn

ro'

6A2

LU.

ul I

LiLU

V.'Id

.0-.

I-V

,421

1 1S

OLUI

Awla-~QA-

V4 04 _

d .~bo

Li

ij ~ 4Ai El '

i-I \-A

I-ILI

~~00

.14

ww

w Z3

.A t

P 1 1: J-4-__-_11

ww

IA

US 0

LUi

UA Ir-

LLd

zz

N 'x

0 w La

w (x

0 ------- 0-...

LLi

Tx-!!

.V-V NOUIVA313 W11'339)

lo

"±/Dg 991 01 031k:21A03 *3SV9 WOOMOWWI

to

z

NMOO49 ION WJV3 l1111,13A

0V71OV3 IOA~ W3d W1 91~3V

1

F w.A.

AIR

AIR WASHER AI

- COLD WATER

II WA5H /WISTTLINGWAiT MNiI TAN

FIG. I

LAY-OUT OF SEMITEC~HNICAL SCALE PLANT

AON & ITIA L

IONF1 DENTIAL

V

F:2

-j0~Lxi

-J

U)

-iU:2

U

1:1IDE

U

400

tI

o w n 1 .. . ..

OLLZ-

a

-7, 1 uImc

F-

L

tA-

17 '1 I-

1

K -

N

4

t

N

'4

- - N

4jN

>5

N

A>

4 A

N>

<A4 1 1

I!4

H* NC~S~fl~D>11 1 A

~>> 4

45